Host: What happens if your boss pays you less than minimum wage or pays you late, or doesn't pay you at all? Data suggests that's not rare in Ontario. Vass Bednar is managing director of the Canadian Shield Institute, and she joins me in studio. How are you doing?

Vass Bednar: I'm good. How are you?

Host: I'm doing well. Let us talk about surveillance pricing. What exactly is it and where do Canadians experience it?

Vass Bednar: Surveillance pricing occurs most of the time when you're shopping online. So in a closed ecosystem, you're either on your computer or you're in an app. It's much harder to comparison shop. And what's happening is that the price you're seeing has been calibrated for you based on a range of inferences that the company is making about you based on personal characteristics, personal data, and some of your behaviour online.

Host: When I think of surveillance pricing, I'm thinking and help me define this. I think of dynamic pricing like an airplane Uber on a very busy, you know, game day. Is that different than...

Vass Bednar: There's some distinctions for sure. There's some distinctions. So I mean, with airlines, I think it's an interesting site to to think about because for a long time we grew comfortable with price fluctuations that were observable, but those were mostly due on due to supply and demand constraints. And then the function of time, right. You're booking far in advance. You might find a better deal. There's some evidence that up to 1000 different data points are now used to calibrate the price that you see when you book your airline. So that's a new dimension, right? That's been supercharged by data. And sometimes a lot of that data has been willingly volunteered by us when we're engaging in loyalty programs that also share our purchase history.

Host: So you're absolutely right that there's distinctions between supply and demand constraints influencing price as well. You know, with food geopolitical pressures, there's a lot that goes into a firm pricing something. But again, increasingly we're seeing with experiments the same people like in the same geography, shopping at the same store online are being served different prices in a kind of price. I'll use the term experiment because it ultimately is an experiment.

Host: A bunch of red flags are going up for me as things that you're saying. So help me understand what personal data specifically are we talking about, how long I'm spending on a website. I recently purchased some dining chairs, and I was very upset about what I saw in terms of the price initially. And then I kept revisiting it over and over and over again, much higher than what it was after I bought it. It went on sale. Is this is this like some of the examples that people are sort of dealing with when it comes to that? What are the sort of triggers of the data points? You mentioned a thousand different things here.

Vass Bednar: But I mean, it could be some of your past purchase history. Have you purchased with that furniture retail multiple times before? Right. Or can they sort of count on, on that purchase? Will that purchase happen quickly? Could be based on them inferring your gender, your race, your level of education, the geography in which you live, and kind of the socioeconomic status of that neighbourhood. Those are some of the elements that are being used. So it's like, it's quite different from how prices varied before. There's always been price variance in markets too. So again, that's not net new, but the how, how firms are able to do this and how granular even discounting can be. Because I'm glad you mentioned, you know, something going on discount later. Sometimes that's because it's like, oh, he's he's coming to this site and gazing at this. He's putting it in his cart and he's not returning. So let's serve him a personalized discount to just sort of tip him over into into making that purchase.

Host: I feel like a bit of a victim. I want to see the chairs. They're lovely. I'll show you after, but I help me understand how can, can, can a consumer know for sure that he or she is being targeted to pay a higher price?

Vass Bednar: It's very difficult. So one, you have to kind of get in the fine print of privacy policies. We also have weak federal privacy laws in Canada that allow and sort of facilitate some of this. The challenge with federal privacy laws that we, through that law, conceive of the associated harms as being individualized. But a lot of the time, the harms that we're experiencing when we're being profiled for pricing cluster us into groups. We are segmented or our our data is de-anonymized, but we're segmented into kind of subsets or groups in terms of where we know it's happening. A lot of the Canadian commentary so far relies on great and recent research that's taken place in the US. So it's being demonstrated by Amazon, by Instacart. Walmart holds patents for this, which is a reminder. Yeah, so some some companies would like to make the ability to gouge us proprietary. And I think that when some of the largest retailers are doing this kind of out in the open and setting a new norm, we should expect that other firms follow.

Vass Bednar: I've looked at some Canadian software companies that purport to do this right. They advertise that they can do this. And when we look at this kind of on the horizon, the potential future of e-commerce Agentic commerce, I don't know if you'd let a chatbot shop for you. Oh, I kind of like shopping when I'm able to do it. But in these systems, it's actually encoded the ability to personalize that kind of price for the individual. So it's an interesting, there's interesting potential in that you think you could code your chatbot to sort of say, find me the best deal on these dining room chairs. But really, that chatbot is also being incentivized by ads to upsell you or tell you that this is the best possible price. And it's kind of not. So it's quite murky.

Host: You don't represent any of these businesses and companies, but I, I'm hoping you can help me understand a little bit. Where are they coming from? Where do they see the benefits to consumers? Or is it really just the bottom line here for them?

Vass Bednar: So it is a practice that's it's not like pay what you can, you know, it is something that's fundamentally about extracting the maximum possible price from people. When I put my sort of efficiency business hat on, I can see the argument that is occasionally made that says, that's our job, right? Our job is to maximize the profit we're able to make, and we should use as much information as is possible to be as calibrated and and granular. When we're experimenting. I understand that, but I also think that discriminating against people based on their personal characteristics is too much and goes too far. You know, the kind of information firms are able to use to, again, I'll say run those pricing experiments because that's what it is. We're we're and like the best subjects, we're not aware that we're being subject to it. They're not in a rush to tell us that they're doing it.

Host: Okay.

Vass Bednar: And I think that says a lot because every time or many times when a firm has been discovered to be doing this. So PlayStation was running an experiment where video gamers who played more often were seeing a higher price for a game. Wow. Do you think people aren't going to share in forums like what price? And they stopped. They said, oh, you know, we were just we're just trying something. Just checking it out. Okay. If you're telling people you're doing that, that's how they respond. It's not a good relationship. And it's kind of a breakdown of trust.

Host: Well, I mean, I can ask a question. I think I know the answer here, not a legal in terms of what they're doing right now. Is that correct?

Vass Bednar: Oh, I mean, look, let's look at Quebec's privacy law 25 says you cannot use people's personal information and data to discriminate against them. Now, that law has not yet been tested in an algorithmic pricing context. If you're someone like me and you read the fine print of the Telecommunications Act, I will say we have precedent in telecommunications that says you cannot charge different people different prices, which makes sense because a lot of price variation started geographically. So when we're all paying for Internet or cell phones, you're not going to upcharge. You know, we actually have a law that says you can't charge people in Rosedale, Toronto more because they can pay. Well, that's not if that's a value that we have as Canadians. That's not something we've sort of been able to have a public conversation about or sort of enshrine in our public policy. And I think we're in this like, interesting.

Host: Yeah. Well, with that being said, we'll go from from Quebec, let's go to Ontario. Doug Ford has said that Ontario will not introduce a ban on surveillance pricing, but one province already has. Manitoba. Tell me, what are they doing and how does it protect consumers?

Vass Bednar: So Manitoba is defining this as a consumer protection problem, which is really interesting. We live in a federation. I was mentioning a federal privacy law pushed that aside. Manitoba is using this lever, and they're saying that it's unacceptable in sort of any context. They're anchoring their work in grocery specifically. Okay. And it'll be interesting to see how they're able to enforce that ban. The tricky thing with bans is they don't necessarily change everything overnight. But I agree with Premier Ford in that I also want a more free and fair market. But I disagree that personalized algorithmic pricing is a free market. When you and I can't really know about it, and we're being deceived and kind of all that consumer surplus is being extracted, that's not a free market. That's just really good deception.

Host: There's also Canada's Competition Bureau as well. So for people who are not familiar independent law enforcement agency, that is to protect and promote competition for consumers, but also businesses with any interest in them sort of watching surveillance and, and sort of policing that.

Vass Bednar: I'm interested. It feels like a bit more of a downstream kind of market policing. But we do have, again, parts of the competition act like double ticketing. You're not supposed to have like two, two competing prices in a particular way. I know that the Competition Bureau also also cares a lot about fake discounting. Like fake discounting is a big way that a lot of us get served these pricing experiments. You know, it says that that dining room chair is $100, but it's actually never retailed for $100. And then all of us are seeing prices underneath it. So look, I'd love to see a private action case taken to the Competition Tribunal and sort of tested there. And I know the Competition Bureau studied it a little bit, probably part of the solution for Canada, but not what's going to revolutionize marketplaces and bring the transparency we need.

Host: Any steps consumers can do to protect themselves.

Vass Bednar: You know, I'm always reluctant to give people advice just because I don't think it's incumbent on us as individuals to sort of fight structural market power, things like one by one. However, there are, you know, with Amazon, there are certain web plugins that give you an alert when a price hits a certain threshold, which I think is also a tacit acknowledgement of the randomized pricing fluctuations people sometimes recommend, including a company in the U.S. that does surveillance pricing. I believe it was Delta Airlines on Twitter advised somebody just use incognito mode. Look, for now, we could all go incognito and sort of hope for the best and not participate in loyalty programs and try to pay with cash where we can, but it's not going to change the core problem. Again, this is currently encoded in e-commerce. It is part of Google's terms of service with advertisers. When they're serving prices. It's hiding in plain sight. And I think we're in such an interesting moment for Canada to to start to react.

Host: Well, we are going to continue to follow this. I really appreciate your time and your insights on this. Thank you so much.

Vass Bednar: Thanks for having me.